Fuel arrives at a forecourt meeting a tight specification and then does something nobody thinks about: it waits. Sometimes for days, sometimes for months, in a steel or composite tank buried under the concrete you are standing on. Almost every folk belief drivers hold about fuel quality — that some brands are cleaner, that you should avoid a station during a delivery, that a quiet forecourt is a bad bet — is really a belief about what happens during that wait. Some of it is right, for reasons more interesting than the folklore.
How clean it is when it arrives
The European diesel standard EN 590 is strict about foreign matter. Water content is capped at 200 mg per kg of fuel — two hundredths of one percent. Total contamination, meaning solid particulate, is capped at 24 mg/kg.
Around those sit the rest of the specification: cetane number minimum 51.0, density 820–845 kg/m³ at 15 °C, sulphur maximum 10.0 mg/kg, FAME up to 7%, viscosity 2.0–4.5 mm²/s, flash point above 55 °C, ash maximum 0.01%, and oxidation stability of at most 25 g/m³ or at least 20 hours depending on the test used.
That last one is the clue. A specification that bothers to define stability over time is a specification written by people who know the product does not get used immediately.
Where the water comes from
Water reaches fuel in storage by three routes, none of which involve anybody doing anything wrong.
Condensation. A partly empty tank contains air, air contains moisture, and the tank walls cool at night. The cycle repeats daily.
Ingress. Rainwater finds its way past fill points and seals, particularly at older installations.
Adsorption from air. The fuel itself takes up moisture from the atmosphere it is in contact with.
Modern diesel is more exposed to the third route than the diesel of thirty years ago, and the reason is the FAME content. Biodiesel is hygroscopic — it readily absorbs moisture from its environment. The same 7% blend that improves lubricity and reduces fossil carbon also makes the fuel more willing to hold water.
Why water at the bottom is worse than water in the fuel: water is denser than diesel, so it separates and settles. That creates a boundary layer where fuel meets water — and that interface is a habitat. It is not a chemistry problem at that point. It is a biology one.
The thing that grows down there
What the trade calls diesel bug is contamination by microbes: bacteria and fungi living in the fuel system.
The organisms are specific and well documented. Bacteria include Clostridium, Desulfotomaculum, Desulfovibrio, Flavobacterium, Acidovorax facilis, Pseudomonas and Sarcina. Fungi include Aspergillus, Candida keroseneae, Fusarium and Hormoconis resinae.
They colonise either the interface between the fuel and the water beneath it, or the tank walls, depending on whether a given organism needs oxygen. What they leave behind is a sludge that blocks filters — and some of these species produce acidic or sulphide by-products that corrode the tank itself.
The timescale is the useful number: left untreated, fuel stays reliable for roughly 6 to 12 months before contamination begins to appear.
Set that against a working forecourt. A busy site turns its tanks over in days. It never approaches six months, and the problem never starts. A site selling a fraction of that volume is operating much closer to the timescale where biology gets a foothold — which is why the meaningful variable is not the brand on the canopy but how fast that particular tank empties.
The folklore, examined honestly
The most repeated piece of forecourt advice is: do not fill while the tanker is delivering, because the delivery stirs up sediment from the bottom of the tank.
The mechanism is plausible on its face. Settled water and particulate do accumulate at the lowest point, and discharging several thousand litres into a tank is not a gentle process. Against that, dispensers draw through filtration, tanks are monitored for water, and modern installations are designed on the assumption that deliveries happen while the site trades.
We could not find authoritative evidence quantifying the risk either way, so we are not going to tell you it is a real hazard or dismiss it as a myth. What can be said is that it is a cheap precaution: if a tanker is on the forecourt and you are not in a hurry, using the next station costs you nothing. Treat it as an inexpensive habit rather than an established fact.
The related belief that some brands sell inherently cleaner fuel is weaker. Fuel meeting EN 590 is fuel meeting EN 590, and neighbouring stations of different brands are frequently supplied from the same terminal. Where brands genuinely differ is in additive packages added at the point of sale, which is a real difference and a much smaller one than the folklore suggests.
What is actually worth doing
- Prefer stations that are busy. Not for the queue — for the turnover. High throughput is the single most effective protection against everything described above, and it costs nothing to choose.
- Be more careful with a station that looks dormant. A forecourt with weeds at the edges and no cars is a different proposition from a quiet one on a busy road. Low volume is the risk factor.
- Take a fuel filter change seriously on an older diesel. It is the component that protects you from everything upstream, and it is the one that fails quietly.
- Do not store diesel for a season and expect it to be fine. Six to twelve months is the published window for untreated fuel. A jerrycan kept "just in case" for two years is not an insurance policy.
- Keep the receipt when something goes wrong immediately after filling. Fuel quality claims are difficult and become impossible without evidence of where and when you bought it.
- Do not over-worry. Contaminated fuel from a normal, trading forecourt is rare. This is a reason to prefer a busy station, not a reason to be anxious at every pump.
The number that would help, and that nobody publishes
Everything above converges on one figure: how quickly a given station sells through its tanks. It is the best single predictor of fuel condition, and it is not published anywhere by anyone. It is commercially sensitive, it is not in any map database, and no regulation requires it.
What exists instead is an imperfect proxy that happens to be visible: whether people are actually using the place. A station with recent driver reports — someone filled here this morning, the queue was short, the price was this — is demonstrably a station that is trading. A forecourt that no passing driver has said anything about for months may be perfectly fine, and it is also the one where the question is worth asking.
That is a modest claim and we would rather make a modest true one than a confident invented one. We cannot tell you a tank's turnover. We can tell you, honestly and with a timestamp, whether anybody has been there lately — and where nobody has, we would rather show you an empty space than a reassuring green dot.